The amount of textual data stored by corporate enterprises has increased significantly in the last decades. A database must provide an efficient way of retrieving relevant information from this data. This task can be formalized as the document retrieval problem. For large text databases a sequential search over the contents is prohibitively slow.